Clinical trials for the treatment of carotid artery disease have focused almost exclusively on a single outcome --prevention of recurrent stroke -- yet accumulating evidence suggests that cognitive function is affected by carotid artery disease and hemodynamic impairment. We propose an ancillary study to the newly-funded, NINDS-sponsored Carotid Occlusion Surgery Study (COSS) to test the hypothesis that cognitive outcomes are relevant and important to measure in the treatment of cerebral hypoperfusion. The COSS trial is designed to test the hypothesis that EC-IC bypass can reduce by 40% the rate of recurrent stroke at 2 years in patients with recently symptomatic internal carotid artery occlusion and Stage II hemodynamic failure as measured by increased oxygen extraction fraction (OEF) on positron emission tomography (PET). We propose to evaluate the cognitive status in randomized COSS patients using global and hemisphere-specific neurocognitive tests in a battery that takes less than an hour to administer. Our specific aim is to determine whether patients who have symptomatic, unilateral carotid artery occlusion and PET evidence of stage II cerebral hemodynamic failure who do not reach a COSS endpoint (ipsilateral stroke) will have better cognitive function 3 months and 2 years following EC-IC bypass compared with an identical group of patients treated with best medical therapy alone. The impetus for our proposal comes from preliminary data in which we demonstrated a quantitative relationship between cerebral blood flow (CBF) and higher cortical function in the setting of carotid artery test occlusion. We and others have further demonstrated cognitive improvement in case series studies of patients with carotid stenosis or occlusion treated with endarterectomy or extracranial-intracranial (EC-IC) bypass. In our proposed study approximately 294 patients will have neurocognitive testing at time of randomization;then 30 days, 3 months, and 2 years after EC-IC bypass or the equivalent time point in the COSS medical arm. All patients will undergo a follow-up PET scan at 3 months to determine actual OEF status. Adjusting for anticipated drop-out due to end-points in the COSS trial, we will have 80% power to detect a 0.5 SD difference in cognitive scores at an alpha level of 0.05. Correlation of cognitive scores with COSS disability and quality of life measures will be made to establish the clinical relevance of change in cognition. In addition to the benefit to the COSS trial if reperfusion surgery can be shown to be improve cognition, positive findings in this study could pave the way for the use of similar cognitive measures in other carotid disease treatment trials.